Kitchenware Image Classification

Kitchenware Image Classification Competition

Participated in the Data Talks Club Kitchenware Classification Competition

Utilized transfer learning, image augmentation, dropout and l2 regularization.

Generated synthetic data using image augmentation to even out class imbalances.

Achieved 0.97726 accuracy on the public leaderboard and 0.98024 accuracy on the private leaderboard


  • TensorFlow
  • EfficientNet
  • Pandas
  • Numpy
  • imgaug
  • Scikit-Learn
  • Jupyter Notebook